报告题目：Backtesting expected shortfall: a duration-severity approach
时 间：2024年05月09日（星期四）11：00
地 点：科研楼18号楼1102
主 办：数学与统计学院、分析数学及应用教育部重点实验室、福建省分析数学及应用重点实验室、统计学与人工智能福建省高校重点实验室、福建省应用数学中心（福建师范大学）
参加对象：感兴趣的老师和学生
报告摘要：We propose an original two-part, duration-severity approach to backtesting Expected Shortfall (ES). Recently, probability integral transform (PIT) based ES backtests have gained popularity, but so far, these tests do not allow to test separately the frequency and severity of the Value-at-Risk (VaR) violations. This latter property is essential, since the ES measures the average loss in case of a Value-at-Risk (VaR) violations. We construct a sequence of inter-violation durations, as well as a sequence of severities corresponding to the PIT in case of a violation. Then we derive orthogonal moment conditions satisfied by these two sequences using the theory of (bivariate) orthogonal polynomials. We propose a simple, model-free Wald test which includes as special cases various unconditional coverage and conditional coverage backtests of both VaR and ES, allowing the backtester to easily identify the mis-specified component(s) of the internal model. The test can also be applied to other systemic risk measures such as the marginal expected shortfall. Simulation experiments suggest that our test has good finite sample properties for realistic sample sizes. Through an application to two stock indices, we illustrate how this approach can be used to analyze the reason of a rejection, when it occurs.
报告人简介：陆扬, 巴黎第九大学(Paris-Dauphine)博士。现任康考迪亚大学数学与统计系副教授。曾任巴黎十三大学讲师。曾主持加拿大自然科学与工程研究理事会（NSERC）, 美国精算师协会(SOA)等科研项目。主要研究方向统计在精算和金融中的应用。主要结果发表在Management Science, Mathematical Finance, Scandinavian Journal of Statistics, Electronic Journal of Statistics, Journal of Time Series Analysis, Journal of Multivariate Analysis, JRSSA, Journal of Banking and Finance, Journal of Applied Econometrics, Insurance: Mathematics and Economics, Scandinavian Actuarial Journal等期刊上。